AFM-IR Depth Sensitivity on Vertically Stratified Polymer Structures

By transducing a sample's photothermal expansion and contraction, the photothermal induced resonance technique (PTIR, also known as AFM-IR) enables characterization of light absorption at scales (≈10 nm) well below the limits imposed by far-field diffraction. However, AFM-IR depth sensitivity on heterogeneous samples remains poorly understood, making quantitative analyses challenging. Here, we examine stratified, three-dimensionally printed, polymer heterostructures engineered to elucidate the AFM-IR signal dependence on detection frequency, light polarization, and measurement modality (i.e., on- or off-resonance). Results show that the signal intensity, derived from absorption in the topmost layer (here, an epoxy-based prism), increases nearly linearly with its thickness regardless of detection frequency or resonance condition, aligning with previous observations in homogeneous (non-stratified) samples. However, the signal from the bottom layer (here, an acrylic photopolymer resin disk beneath the epoxy prism) can be amplified, attenuated, or remain unchanged, depending on the detection frequency and measurement modality. On-resonance, the signal coming from absorption in the bottom, acrylic layer always decays approximately exponentially with the thickness of the top, epoxy layer, with a frequency-dependent decay rate. Notably, absorption from the acrylic layer was detected even beneath the thickest epoxy region (here, ≈440 nm) for all measurement conditions, highlighting the considerable subsurface sensitivity of AFM-IR. We propose adopting a ≈95% signal decay as a convention to describe a practical AFM-IR measurement depth sensitivity (probed depth), as opposed to the previously suggested ≈63% decay. This work advances AFM-IR-based nanoscale chemical imaging by elucidating the relationship between signal attenuation and sample stratification, which represents a critical step toward more-quantitative and three-dimensional compositional analyses.